Prognostic Impact of Clinical and Molecular Markers in Normal Karyotype AML-Results from the AMLCG 2000 Trial.

Background: Prognosis of AML is influenced by different clinical and molecular alterations. We performed a multivariate analysis including five molecular markers NPM1, FLT-ITD, CEBPA, FLT-TKD and MLL-PTD combined with clinical parameters at initial diagnosis to refine risk stratification.

Patients and methods: Prognostic impact of clinical and molecular parameters in respect to OS, EFS, RFS and CR was assessed in 803 patients with normal karyotype included in the AMLCG (German AML Cooperative Group) 2000 trial until 01/2006. Patients were randomly assigned to treatment with TAD (thioguanine, conventional-dose AraC, daunorubicin) followed by HAM (high-dose AraC, mitoxantrone) or with the double-induction regimen consisting of two courses of HAM (quotation Buechner JCO 2006). Patient age ranged from 17 to 85 years (median: 60 yrs). 51% of patients were male, 49% female. 81% of patients had de novo AML. Performance status was normal or slightly impaired in the majority of patients (71% ECOG 0/1). Median blood counts at diagnosis were: Hb: 9.2 g/dl (4.2–16.4 g/dl); WBC: 16.0 G/l (0.1–798.2 G/l); platelets: 58 G/l (0.02–643 G/l), LDH: 410 U/l (8–14332 U/l) and bone marrow (BM) blasts: 80% (6–100%). Molecular markers’ mutation status and all mentioned clinical parameters were included in univariate analyses. In multivariate analyses only univariate significant parameters were used.

Results: In 560 patients with all five molecular markers analyzed by routine molecular techniques at diagnosis the frequency of mutations were the following: 52.7% NPM1+, 29.3% FLT3-ITD+, 6.1% FLT3-TKD+, 7.5% MLL-PTD+ and 7.5% CEBPA+. The majority of analyzed patients (44.1%) showed one single mutation only. About one quarter of patients displayed either none (27.5%) or two (26.2%) mutations. A minority of 2.1% had 3 mutations, whereas the combination of four or all five molecular alterations was not found. The most frequent single mutation was NPM1 (28.4%), followed by FLT3-ITD (5.4%), CEBPA (4.8%), MLL-PTD (4.6%) and FLT3-TKD (0.9%). The combination of FLT3-ITD and NPM1 was detected in 18.8% of patients. Complete remission (CR) rate was 65.1%. Median overall survival (OS), event-free survival (EFS) and relapse-free survival (RFS) were 19.3, 7.7 and 17.2 months. Multivariate analyses identified the following parameters to have significant impact on prognosis. OS: NPM1, FLT3-ITD, WBC, age (p<0.0001 each) and CEBPA (p=0.003); EFS/RFS: NPM1, FLT3-ITD and age (p<0.0001 each / p<0.0001 each) and LDH (p=0.020 / p=0.040); CR: NPM1 and age (p=0.001 each).

Conclusions: Our data show in a large cohort of 560 patients that at least one molecular marker can be identified in 72.5% of patients with NK-AML. The NPM1 mutation and age are the only parameters with an independent impact on all outcome parameters (OS, EFS, RFS, CR). These data provide the basis for a prognostic model in NK-AML that can be used for risk stratification and selection of patients that will benefit from allogeneic stem cell transplantation.